In a recirculation system, five functions must occur to provide life support for the fish. There must be: 1) a place for the fish to live (fish tank), 2) a way of moving the water (pump), 3) a way of removing solid waste (e.g. bead filter), 4) a method of nitrification (e.g. trickle filter), and 5) a method of gas exchange (e.g. aeration). These functions are often accomplished by individual components as suggested above, but in some cases one component may provide two functions. For example, a bead filter removes solid waste and also provides for some nitrification. In other cases, a single function may be carried out by two or more components. One example, would be a system like the one shown below that uses a bead filter to remove large to moderate size solid waste particles and a foam fractionator to remove very fine particles and oils. Another would be a system that uses some type of pure oxygen system to add oxygen and then requires a gas stripping column to remove carbon dioxide.
System components can be made of a variety of materials, but they must be inert and not reactive with water. Galvanized and copper pipe are suspect, because under conditions of low hardness and low pH, the zinc or copper ions can accumulate in the water causing toxicity to fish. Concrete, either poured or in the form of block, is widely used in traditional aquaculture for raceways. In recirculation aquaculture, care must be used with concrete because in the absence of continuous fresh water the concrete can leach compounds that can raise pH too high. Old, or "cured" concrete is usually all right and it can also be painted with epoxy paint. Two part epoxy paints are expensive but very useful for creating a durable, non-reactive surface that can be disinfected. It can be used on a variety of surfaces including wood and metal.
There are two approaches to scaling systems to grow more fish. One is to build a bigger system with larger components. The other is to build multiple units of a more modest size system. Both approaches have been used in commercial recirculation ventures. The single large system may have some advantages of scale, that is, it may be cheaper to build on a per-gallon basis, but it lacks the flexibility and redundancy of smaller, multiple units. A pump failure or a disease outbreak in a single large system may cause the loss of all the fish being grown. In a multiple unit system, however, only a fraction of the fish would be lost. Also, with multiple units different species or different sizes of fish may be grown that would be difficult or impossible if they were mixed together. Smaller systems are easier to harvest, as well.
In a functioning system, all of the components must work together. Like any complex system a recirculation system is only as robust as its weakest link. At any given time there is always one limiting factor to increased fish production. An efficiently designed system is made of components of approximately equal capacity. There is no point, for example, in having nitrification capacity far in excess of what the gas exchange can support. How does one design such a system?
The bad news for a biologist, like myself, is that the traditional way of designing efficient systems is an engineering process called mass balance. What this entails is calculating the size of each component based on projected load and known relationships. You begin with the amount of fish you want to grow. This allows you to calculate how much food will need to be fed, which in turn allows you to calculate the amount of oxygen required to metabolize that feed. Other equations predict the amount of oxygen that a given aeration system can move, etc. This exercise is continued until a system design is devised that theoretically supports the desired fish load without expensive overcapacity in any area. This process is not perfect. There are areas where the required supporting information is either not present or imperfect, thus real world adjustments and judgement calls will have to be made. Nonetheless, if a sizeable investment is going to be made in a recirculation system, the first expense should be an engineering design based on mass balance. For an excellent treatise on the use of mass balance in recirculation aquaculture, see Timmons et al. "Recirculating Aquaculture Systems" published by the USDA Northeastern Regional Aquaculture Center (ISBN 0-9712646-0-0). The fact is, successful recirculation aquaculture is based as much on engineering than biology and much academic research in this area is conducted in agricultural engineering departments, not fisheries departments.
It is possible to design a smaller, reasonably efficient system without a systematic mass balance analysis, but the same principle is involved. You begin with the weight of fish you wish to support and then estimate the size and type of each component needed. Instead of using equations you would use rules of thumb and the practical experience of others. A system constructed by this method would likely need more adjustments than one developed through rigorous calculation. Visiting operating recirculation systems is invaluable to someone developing a system design.